An Investigation into Hospital Lengths of Stay and Costly Readmissions Objectives Readmissions to the hospital within a short period after discharge are often indicators of poor quality of antecedent care, yet are frequently preventable. Readmissions are also prevalent among U.S. hospitals and are significant cost drivers. A potential explanation for high readmission rates that previously has not been explored is suboptimal length of stay in the index hospitalization. If in response to financial incentives of payers or providers, patients are discharged too soon, then many may need to return to the hospital for additional inpatient care within a short period of time. This project aims to evaluate the relationship between length of stay and readmission, and to assess potential cost savings associated with improving the balance between longer initial hospital stays and reduced readmissions by carrying out the following aims: Aim 1: Examine the extent to which the length of hospital stay is associated with the likelihood of inpatient hospital readmission. Aim 2: Evaluate the cost trade-off between hospital length of stay and readmissions. Methods The primary data source is the 2008 HCUP State Inpatient Database for New York State. Logit models will be used to explore the relationship between length of stay in an index hospitalization and probability of readmission within 30 days of discharge, controlling for predisposing patient characteristics as well as other factors potentially associated with readmission prevention. Two-part models will link the results of the probability models to models of readmission cost estimated using generalized linear modeling approaches, in order to generate values of expected readmission costs. These will be compared with estimated values of the marginal costs associated with keeping patients longer during the initial hospital stay. Subgroup analyses for selected conditions with relatively high readmission rates and/or high costs will be conducted. PUBLIC HEALTH RELEVANCE: An Investigation into Hospital Lengths of Stay and Costly Readmissions Relevance. This project will build insights concerning the potential relationship between length of stay in the hospital and likelihood of hospital readmission. It will also expand knowledge regarding the cost implications of potentially preventable readmissions. This presents an outstanding opportunity to improve value in health care delivery by building insights that will contribute to containing hospital costs simultaneously with improving hospital quality of care through reduced readmissions.